Applied Business Analytics for Marketing Decision-Making

Business Analytics and Data Visualization

Dr. Jose Mendoza

Academic Director and Clinical Associate Professor

MS in Integrated Marketing

New York University, School of Professional Studies

Fall 2026

Version 1.0 · July 2026

Except where otherwise noted, this guide is licensed under CC BY 4.0.

Copyright

Applied Business Analytics for Marketing Decision-Making: Business Analytics and Data Visualization

Copyright and License

Copyright © 2026 Jose Mendoza. Except where otherwise noted, this work is licensed under the Creative Commons Attribution 4.0 International License (CC BY 4.0). You are free to share and adapt the material for any purpose, provided you give appropriate credit, link to the license, and indicate whether changes were made. The full license text is available at https://creativecommons.org/licenses/by/4.0/.

Third-party material — including quoted or adapted definitions, figures, screenshots, and datasets attributed to their original sources — remains the property of the respective rights holders and is used here with attribution. The CC BY 4.0 license does not extend to that material, to the New York University name and marks, or to the third-party software interfaces shown in screenshots.

Prepared For

This course material and reading guide was prepared for instructional use in Business Analytics and Data Visualization (INTG1-GC 2300) in the MS in Integrated Marketing program at the New York University School of Professional Studies.

Distribution

The guide is distributed to enrolled students. Datasets, notebooks, templates, and project files are distributed through the companion repository. Because the guide is released under CC BY 4.0, students, instructors, and other readers may retain, reuse, and adapt it beyond this course, with attribution.

Suggested Citation

Mendoza, J. (2026). Applied business analytics for marketing decision-making: Business analytics and data visualization (Version 1.0). New York University School of Professional Studies.

Trademarks

Product and company names mentioned in this guide (including ChatGPT, Claude, Gemini, GitHub Copilot, Google Colab, NotebookLM, and Tableau) are trademarks of their respective owners and are used for identification purposes only. Their mention does not imply endorsement. StyleCraft Collective, the running case used throughout the guide, is a fictional company created for instruction.

Companion Repository

Datasets, notebooks, templates, and figure sources for every chapter: https://github.com/jrmst102/businessanalytics

Version and Date

Version 1.0 · July 2026 · Language: English (United States)

Dedication

To my students, whose questions continue to show that analytics is not only about finding answers, but about learning how to ask better questions.

Preface

Business analytics is changing.

For many years, learning analytics meant learning to perform a sequence of technical tasks: clean a dataset, run a model, create a chart, and interpret the output. Those tasks still matter. Students still need to understand data, measurement, relationships, models, visualizations, and recommendations. But the way analysts perform this work has changed.

Today, business analysts increasingly work with AI assistants. Tools such as ChatGPT, Claude, Gemini, NotebookLM, GitHub Copilot, Colab AI features, Tableau AI capabilities, and other emerging platforms can generate code, explain errors, summarize data, suggest analyses, draft interpretations, and create first versions of charts and dashboards. These tools are becoming part of everyday professional work.

But faster output does not automatically produce better analysis.

AI tools can be useful, but they can also be wrong, vague, overconfident, or misleading. They may produce code that runs but answers the wrong question. They may generate charts that look polished but distort the message. They may summarize findings in language that sounds convincing but goes beyond what the data supports. In an AI-assisted environment, the analyst’s responsibility does not disappear. It becomes more important.

This guide is written for that environment.

The goal is not to turn every student into a software engineer or statistician. The goal is to help future marketing and business professionals become capable, critical, and responsible users of analytics. Students will learn how to prepare data, explore relationships, build preliminary models, evaluate evidence, create visualizations, and communicate recommendations. They will also learn how to work with AI tools without surrendering judgment to them.

Concept

What an Analyst Does Now

Modern marketing analysts do not merely produce code, charts, models, or dashboards. They frame problems, supervise AI-assisted workflows, verify outputs, and translate evidence into decisions.

Throughout this guide, students will use AI assistants as part of the analytics process. But AI use is not treated as a shortcut. It is treated as a professional workflow that requires discipline. Students must specify what they are trying to do, predict what should happen, verify what was produced, explain the result, and document the process.

Analytics becomes valuable only when it improves decision-making. That principle guides every chapter.

Who This Guide Is For

This guide is designed for graduate students in marketing, integrated marketing, business analytics, and related professional programs. It assumes that students may have limited prior experience with programming or statistics, but that they are preparing to make or support decisions in data-rich business environments.

The guide is especially appropriate for students who need to:

• Use data to support marketing decisions.
• Prepare, clean, and explore real-world datasets.
• Understand customer, campaign, and business metrics.
• Analyze relationships among marketing variables.
• Build preliminary predictive models.
• Evaluate model output and communicate uncertainty.
• Create analytically supported visualizations.
• Use AI assistants responsibly in analytics work.
• Translate findings into managerial recommendations.

The emphasis is practical, but not superficial. Students will learn technical concepts because those concepts matter for interpretation. They will use tools because tools make analysis possible. But the final goal is not a completed notebook, a model score, or a dashboard. The final goal is a recommendation that helps a decision-maker act with better evidence.

How This Guide Is Organized

This guide is organized into three parts and thirteen chapters, supported by six appendices and a back matter section. Chapters are assigned one per week and align with the session sequence in the course syllabus, so the reading for each session is the working material for that session.

Part I: Foundations of Marketing Analytics

Part I introduces analytics as a form of decision support. Students learn how to frame marketing problems, define variables, prepare data, explore datasets, and summarize customer and business patterns. These chapters establish the foundation for later modeling and visualization work.

Part II: Predictive Marketing Analytics

Part II focuses on methods used to understand relationships, segment customers, forecast outcomes, evaluate experiments, and build predictive models. The emphasis is not only on applying methods, but also on understanding what those methods mean, when they are useful, and how they can be misinterpreted.

Part III: Data Visualization and Analytics Communication

Part III focuses on using visualization to analyze data and communicate findings. Students learn how to select appropriate visual forms, avoid misleading designs, create dashboards, and turn analytical findings into executive-facing recommendations.

How Each Chapter Works

This guide is designed to be used one chapter at a time. Each chapter functions as a weekly learning unit and follows a consistent structure:

Marketing Decision Context. Each chapter begins with a practical marketing problem, connecting the technical topic to a decision that a manager, analyst, or marketing team might face.

Key Concepts. Each chapter introduces the core ideas needed to understand the method or workflow. Definitions are explained in plain language and tied to marketing examples.

Analytics Workflow. Each chapter walks through a practical analytic process. Depending on the topic, this may involve preparing data, summarizing a dataset, estimating a model, evaluating output, or designing a visualization.

AI-Augmented Analytics Lab. Each chapter includes an AI-assisted activity. Students use an AI tool to support part of the workflow, such as generating starter code, proposing chart types, explaining output, debugging an error, or drafting a managerial summary.

Verification Check. Each chapter includes checkpoints that ask students to verify the work. These checks may involve row counts, summary statistics, missing values, model diagnostics, chart scales, benchmark comparisons, or interpretation limits.

Business Analytics in Practice. Each chapter closes the loop to industry with short vignettes drawn from real analytics practice, ending with what the chapter’s skill looks like in a first analyst job.

Managerial Interpretation. Each chapter asks students to translate analytic output into business language. The focus is on what the result means, why it matters, and what action it may support.

Reflection. Each chapter ends with reflection questions that help students connect the technique to professional practice, ethical judgment, and decision-making.

Companion Site and Course Files

The files and code used in this guide are available through the companion GitHub repository: https://github.com/jrmst102/businessanalytics

The companion site provides working materials for the weekly chapters, including Python notebooks, example datasets, code templates, verification exercises, project starter files, Tableau support files, and AI-use documentation templates when applicable.

Students should use the companion site alongside the assigned chapter. The guide explains the concepts and workflow; the repository provides the files needed to practice and complete the exercises. The repository is not a substitute for the chapter readings, and students should not submit repository files without adapting, explaining, and verifying them. The purpose of the companion site is to support analytic practice, not to provide finished answers.

Recommended Companion Site Workflow

• Read the assigned chapter.
• Open the corresponding notebook, dataset, Tableau file, or template.
• Review the business question and data.
• Specify the analytic task before prompting an AI assistant.
• Run or adapt the code or visualization workflow.
• Verify the output.
• Interpret the result in managerial language.
• Complete the assignment and AI-use documentation.

The AI-Augmented Analytics Workflow

This guide uses a four-part workflow: specify, predict-then-verify, explain, and document. Chapter 1 introduces the workflow; every chapter afterward applies it and gives it a chapter-specific verification theme.

1. Specify

Before using an AI assistant, define the analytic task. A good specification identifies the business question, the unit of analysis, the relevant variables, the intended method, and the criteria for success. AI tools respond to the quality of the instructions they receive: a vague prompt often produces plausible-looking but unfocused work, while a precise specification improves the chance that the tool will produce something useful. Ask:

• What decision is being supported?
• What question must the analysis answer?
• What is the unit of analysis?
• Which variables are needed?
• What method is appropriate?
• What would a useful result look like?

2. Predict, Then Verify

Before running AI-generated code or accepting AI-generated output, predict what should happen. After running it, verify that the result makes sense. Verification is one of the most important habits in analytics: it protects against hidden mistakes, incorrect assumptions, and polished but inaccurate output. Common checks include:

• Row and column counts.
• Variable names and data types.
• Missing values and duplicate records.
• Summary statistics and known totals.
• Chart scales and labels.
• Model fit and performance metrics.
• Holdout results or baseline comparisons.
• Whether the conclusion follows from the evidence.

3. Explain

Students must be able to explain what each step does and why it matters. This includes code, transformations, visualizations, models, and recommendations. Being able to run an analysis is not the same as understanding it. Ask:

• What does this step do?
• Why is this method appropriate?
• What does the output mean?
• What does the output not mean?
• What should a marketing manager take away?

4. Document

Every major assignment and project requires documentation of AI use. Documentation records which tools were used, what prompts were submitted, what the tool produced, what was accepted or rejected, and how the final work was verified. Documentation is not an administrative formality; it is part of analytic accountability. The AI-Use Documentation Template in Appendix D asks:

• Which tool did I use?
• What did I ask it to do?
• What did it produce?
• What did I change?
• What did it get wrong?
• How did I verify the final result?
• What judgment did I apply?

Responsible Use of AI

AI assistants are expected tools in this guide, but they are not substitutes for learning.

Students may use AI tools to:

• Generate starter code.
• Debug errors.
• Explain Python, Tableau, or analytics concepts.
• Suggest analytic approaches.
• Draft preliminary chart ideas.
• Summarize source material.
• Improve the clarity of written explanations.
• Create first drafts of visualizations or dashboards.
• Translate technical findings into managerial language.

Students may not use AI tools to avoid understanding the work. The following practices are unacceptable:

• Submitting AI-generated work that cannot be explained.
• Submitting AI-generated code without checking whether it works correctly.
• Presenting AI-generated interpretation as one’s own reasoning without review.
• Fabricating verification steps.
• Uploading confidential, proprietary, or personally identifiable data to external AI tools.
• Using AI tools during assessments designated as AI-free.
• Omitting required AI-use documentation.

Concept

The Analyst of Record

AI may assist the analysis, but the student remains the analyst of record.

Required Tools

Google Colab. Google Colab is used for Python-based analytics. Students will use it to load data, clean datasets, summarize variables, create exploratory visualizations, build models, and evaluate results.

Tableau. Tableau is used for visual analytics, dashboard design, and executive-facing communication. Students will use Tableau Desktop and Tableau Prep in the visualization portion of the course and in the final visualization project.

AI assistants. Students are expected to use at least one general-purpose AI assistant and one code-capable assistant. Examples include ChatGPT, Claude, Gemini, GitHub Copilot, Colab AI features, and similar tools approved by the instructor.

Companion repository. The companion repository at https://github.com/jrmst102/businessanalytics provides files, notebooks, datasets, templates, and project materials used with this guide.

Course materials. Datasets, Colab notebooks, Tableau files, project templates, rubrics, and AI-use documentation templates are distributed through the course learning management system (Brightspace) and, where appropriate, through the companion repository.

What Students Are Expected to Learn

By the end of this guide, students should be able to:

• Prepare data for forecasting sales and marketing outcomes.
• Analyze information to discover and explore meaningful relationships in datasets.
• Develop strategic recommendations based on the interpretation of real-world datasets.
• Produce analytically supported data visualizations to communicate sales and marketing opportunities.
• Use AI assistants responsibly to support, verify, and communicate analytics work.

The first four outcomes are the course learning outcomes stated in the syllabus. The fifth does not replace them. It describes the environment in which the first four now operate.

How to Read This Guide

This guide should be read actively.

Do not treat the chapters as background readings to skim before class. Each chapter is a working session. Read with the dataset, notebook, or tool open. When a concept is introduced, test it. When code is shown, run it. When a prompt is suggested, revise it. When a chart is produced, critique it. When a recommendation is written, ask whether it follows from the evidence.

A useful reading routine is:

• Identify the business problem.
• Review the key concepts.
• Work through the example or lab.
• Use an AI assistant to extend, explain, or troubleshoot the work.
• Verify the result manually.
• Translate the finding into managerial language.
• Complete the reflection or documentation activity.

The goal is not to memorize every command. The goal is to develop analytic judgment.

A Note on Code

This guide uses Python, but it does not assume that students already know how to code.

The purpose of code in this guide is to support analysis. Students should understand what the code does, how to run it, how to recognize when it fails, and how to interpret the output. AI assistants can help write or debug code, but students remain responsible for the result.

When working with code, focus on four questions:

• What is the input?
• What transformation, calculation, or model is being applied?
• What is the output?
• How do I know the output is correct?

This approach is more important than memorizing syntax.

A Note on Visualization

Data visualization is not decoration. It is a form of reasoning.

A good visualization helps the viewer see a pattern, compare values, understand a relationship, identify an exception, or make a decision. A poor visualization can obscure the truth, exaggerate a finding, or create false confidence.

This guide treats visualization in two ways. First, visualization is used for analysis: exploring data, finding patterns, and checking assumptions. Second, visualization is used for communication: presenting findings to decision-makers through charts, dashboards, and visual reports.

AI tools can now generate charts quickly. That makes visual judgment more important, not less. Students must learn to evaluate chart type, scale, labeling, hierarchy, color, annotation, and the connection between the visual and the business question.

A Note on Managerial Recommendations

The final step in business analytics is not the model, the chart, or the dashboard. The final step is the recommendation.

A recommendation should be clear, evidence-based, actionable, and appropriately qualified. It should tell the decision-maker what the analysis suggests, why it matters, what action should be considered, and what uncertainty remains.

A strong recommendation answers:

• What did we find?
• Why does it matter?
• What should the organization do?
• What are the risks or limitations?
• What should be measured next?

Analytics becomes valuable when it improves decisions.

Acknowledgments

This guide grew out of teaching Business Analytics and Data Visualization to graduate students in the MS in Integrated Marketing program at the New York University School of Professional Studies.

I am grateful to the students whose questions, projects, challenges, and professional experiences shaped the design of this material. Their work made clear that analytics education must evolve with the tools now used in practice while preserving the core disciplines of reasoning, verification, and communication.

I also acknowledge the broader community of educators, analysts, researchers, and practitioners who continue to rethink how analytics should be taught in an AI-assisted world.